UNIVERSE" - ■JJNOIS i &t KL URBANA-CHAMPWGN UROLOGY sr CD =2 FIELDIANA / Geology Published by Field Museum of Natural History Volume 33, No. 23 April 4, 1977 This volume is dedicated to Dr. Rainer Zangerl The Stature and Weight of Sterkfontein 14, a Gracile Australopithecine from Transvaal, As Determined from the Innominate Bone Charles a. reed DEPARTMENT OF ANTHROPOLOGY UNIVERSITY OF ILLINOIS AT CHICAGO CIRCLE AND research associate Field museum of natural history AND DEAN FALK Department of Behavioral Science Rollins College, Winter Park, Fla. With the first announcement of Australopithecus, a new genus of hominoid primate from South Africa, by Raymond Dart in 1925 speculation began concerning the size (height and weight) of the individual. These speculations increased after Robert Broom dis- covered additional examples of the genus in the 1930's and after. With the distinguishing of two populations, one of presumably smaller individuals (gracile) and the other of presumably larger individuals (robust) by Broom in the 1930's and 1940's (Broom, 1950) and by many paleo-anthropologists since, and with the increasing certainty that the australopithecines were members of the human family, Hominidae, the question of size has persisted. Of the two components of "size" (height and weight), that of height is primary and most research on the topic has focused on this aspect. The remains of australopithecines have been found in eastern and southern Africa dating between 5,500,000 and 1,000,000 years ago, but are usually scattered and mostly broken; predation by large carnivores seems to have been the major cause of death. Intact Library of Congress Catalog Card No.: 76-48946 The LiteTHrv dl^tw Publication 1252 423 jUN 06 1977 424 FIELDIANA: GEOLOGY, VOLUME 33 skeletons are unknown, and indeed the long bones of the appen- dages are almost always broken and with parts missing. Thus the known relationships between the lengths of the long bones in a modern individual and the stature of that individual (Trotter and Gleser, 1952, 1958) cannot be used directly on australopithecines. Typically, various means of estimating the lengths of broken long bones have been attempted, and then the height of the individual calculated, or a measure of a width of some surviving part of a long bone has been determined, and a ratio established on modern skeletons between this width and the length of the bone has been used to determine the probable length of the fossil bone, from which datum the stature of the individual has been estimated, within a probable range. The latest, most complete, and most satisfactory effort of this kind has been accomplished by McHenry ( 1974), whose publication can be consulted for descriptions of techniques and a review of previous investigations. One of the specimens used variously by several investigators in their attempts to determine heights of australopithecines is a broken femur, missing the whole of the distal end as well as the head and part of the neck, of a partial skeleton (Sts 14) collected at Sterkfontein, Transvaal, South Africa, by Broom and Robinson (1947). Sts 14, a gracile australopithecine,1 consists (in addition to the partial femur) of parts of the vertebral column and ribs, and of a nearly intact right innominate and some fragments of the left one. This pelvis, reconstructed mostly from the near-intact right half, has been studied extensively with regard to problems of locomotion in the gracile australopithecines (Robinson, 1972) but seemingly no one has attempted to use it to determine the height of the living individual who existed in South Africa some two or three million years ago.2 'The taxon of the gracile australopithecine hominids is usually considered to be Australopithecus africanus Dart 1925, but Robinson (1967, 1972), thinking in terms of clades and arguing that these gracile forms were directly ancestral to Homo erec- tus and thus to living humans, has called them Homo africanus, a practice followed by only a minority of anthropologists. We shall here avoid the taxonomic problem by simply calling them gracile australopithecines. 2No absolute dates are available for any of the five sites in southern Africa from which australopithecines have been collected. For Sterkfontein, estimates of the time of deposit of the faunal remains (including hominids) has ranged from 3,000,000 to 800,000 years ago (see Tobias, 1973, and Sampson, 1974, for discussions); the present conclusion seems to be that 2,500,000 may be the best estimate for Sterk- fontein, although deposition may well have continued for 100,000 years or more (Tobias, 1973, p. 318). REED & FALK: STERKFONTEIN 14 425 The present study is based upon the casts (pi. I) of right innomi- nates of Sts 14 and of a female Mbuti pygmy (MMC 18) from the osteological collection of the Makerere Medical College, Uganda, as made by the Wenner-Gren Foundation for Anthropological Re- search. The errors inherent in reaching conclusions based on measurements taken from casts have been emphasized recently, particularly by Clarke and Howell (1972), but the casts made by the Wenner-Gren Foundation have been produced in a hard plastic with extreme care with respect to similarity to the original, for the particular purpose of use by the majority of paleo-anthropologists to whom the original specimens in Africa are not available. While measurements on such small objects as teeth might yield an unduly high percentage of error between cast and original, we do not believe that errors in measurements on as gross an object as an innominate and its cast will be sufficient to change the final conclusions, as based upon a technique which unavoidably incorporates other, and probably greater, chances of error. The pubis, particularly, but also the inferior ramus of the ischium of Sts 14 (pi. I) are in part broken and perhaps somewhat warped, so that measurements involving length or breadth of these parts of the innominate cannot be regarded as trustworthy. The top of the ace- tabulum at the anterior inferior iliac spine is also indefinite and the anterior superior iliac spine and the adjacent part of the ilium are missing in the original, as is a minor part of the iliac crest, so that good comparable measurements, insofar as we can determine from inspection of the casts of the two innominates, are limited to two, marked AB and AC on the outline drawing (fig. 1). Table 1. Measurements (in cm.), and ratios of measurements, on right innomi- nates of MMC 18 and Sts 14. Ratio AB AC AC/AB Mbuti pygmy MMC 18 12.13 7.20 59.36% Gracile australopithecine Sts 14 12.09 7.42 61.37% Ratio, STS 14/MMC 18 99.67% 103% The remarkable similarity in size1 between the innominates of the 'This similarity was first noted by Professor Alexander Galloway when he was teaching at the Medical College of Makerere, Uganda. In 1950 he showed MMC 18 to Raymond Dart, who then took the innominates with him on a visit to the U.S.A. Thus the choice by the Wenner Gren Foundation of the innominate of MMC 18 as a model from which to make a cast for comparison with the innominate of Sts 14 was purposeful, not fortuitous, based upon the remarkable similarity of the two, a simi- larity first noted by Galloway. 426 FIELDIANA: GEOLOGY, VOLUME 33 Fig. 1. Outline drawing of a right innominate, human, showing the measurements AB and AC used in comparing Sts 14 and MMC 18. Mbuti pygmy and Sts 14 is apparent from the near-identity of the measured lengths and ratios recorded in the first two columns of T*.ble 1. The third column, the ratio AC/AB, is introduced as fur- ther evidence of the similarity of the relationship between measure- ments of AB and AC in the two specimens. Several investigators have previously noted the similarity of size between the innominate of the Mbuti pygmy and that of Sts 14, without presenting the measurements and ratios of Table 1, but seemingly no one has thought to use this similarity, particularly of the ilium, as a basis for comparison of stature, probably because the height of the Mbuti pygmy has remained unknown. It was our in- quiry on this point to the Wenner-Gren Foundation that led to our learning that the cast was based on skeleton no. 18, in the anatomi- cal collection of the Makerere Medical College (MMC), Kampala, O C8 w o - C .» o -4J O CO _ a s § to fe DO § — u .s ^ £ i o >-* i^ 9 >» a * 1| bo o of -c C « O CO .9 o c — ' i-S 5 a 2 ° boPQ CO cu o fe 2» o —• ~S >?m 427 Right Left 35.10 35.10 34.75 34.65 31.20 30.65 428 FIELDIANA: GEOLOGY, VOLUME 33 Uganda. A second letter from us, directed to that institution, in- cluded drawings of a femur and tibia, with indications on the draw- ings of maximum and bicondylar lengths of the femur and of maxi- mum length of tibia. Dr. J. M. Kiggundu very kindly made the measurements on the original skeleton and returned our drawings to us, with the measurements (table 2) and the additional informa- tion that the specimen had been a female no less than 20 years old,1 but that the stature was not recorded. Table 2. Measurements of femur and tibia of female pygmy, MMC 18 (cm.), as made by Dr. J. M. Kiggundu. Femur, maximum length Femur, bicondylar length Tibia, maximum length The difference between the measured lengths of the right and left tibiae have led us to place less reliance on these measurements than on those of the femora; furthermore, since calculations of stature on the bases of femoral maximum length and bicondylar length yield results not significantly different (Trotter and Gleser, 1958), we present below only our conclusions concerning the stature of the Mbuti pygmy as determined on the basis of the femoral maximum length.2 When we determined the height of MMC 18 from our estimate of her femoral length by use of the standard formula of Trotter and Gleser (1952, table 13, and table 13, Appendix 4) for American Negro females, we derived a mean figure of 140.8 cm. (4 ft. 7% in.). Such a stature, as calculated, is presumably too great, as the regres- sion formula (2.28 X Maximum Femoral Length +59.76 ±3.41 cm) used by Trotter and Gleser (1952) for derivation of height from femoral length is based on individuals of normal size and yields 'Dr. Kubet Luchterhand of Roosevelt University, Chicago, who made the mold from which the cast of MMC 18 was produced, has told Reed (personal communica- tion) that, at the time he was working with the original innominate, it was examined by Prof. Adolph Schultz, then of John Hopkins University, who estimated the age of the individual to have been 19 years. 2After we had arrived at this stage in our investigations, we discovered that skele- ton MMC 18 had already been the object of an osteometric study (Toerien, 1954). The published figure of 34.9 cm. for maximum length of the (presumably) right femur is in excellent agreement with that of 35.1 cm. determined for us by Dr. Kig- gundu. We split the difference and for the purposes of this paper assume that the maximum length of the femur of MMC 18 was 35.0 cm. REED & FALK: STERKFONTEIN 14 429 erroneously high results for extremely short individuals such as pygmies. This is due in minor part to the disproportionate value of the added constant relative to the short stature but in major part simply to the use of an equation based on a normal population to obtain an estimate for an individual from an inappropriate popula- tion—i.e., an African pygmy whose height (see below, next para- graph) is within the middle range for that population (Schebesta, 1938, p. 231, graph 1). In any case, we think that Robinson (1972, pp. 231-232) was misled by the tables in Trotter and Gleser (1952) into thinking that he could validly use their regression formula for an estimate of the height of an individual as short as Sts 14, as based upon his estimate of her femoral length. Indeed, there are extensive discussions in the literature on the error of using a regres- sion equation based on one population to estimate the height of an individual, as based on the lengths of the latter 's long bones, when that individual is from another population. We mention Allbrook (1961), with his useful bibliography, as one example of such a dis- cussion. Approaching this problem from another viewpoint, we note for the 177 American black females in the Terry Collection that the mean length for all femora is 27.2 per cent of the mean stature of those individuals (Trotter and Gleser, 1952, table 5), and also find that in the Mbuti female for which there exists a published figure (Flower, 1888)1 for both stature (123.1 cm. = 4 ft. % in.) and fe- moral length (33.4 cm.), the ratio is 27.1 per cent. This delightful agreement of femoral-statural ratios, derived from two such dispar- ate sources, led to our calculation of the height of MMC 18 as close to 128.9 cm. (4 ft. 2% in.), as based on her femoral length of 35.0 cm.2 'Flower wrote only of the Akka (more usually spelled Aka). The Aka, however, with the Efe and the Basua, comprise the pygmies of the Ituri forest, otherwise known collectively under the names of Mambuti, Mbuti, Bambuti, or Eastern Twiden. 2If, for the sake of completeness, we use the more involved method of Pearson ( 1899, p. 230 and PI. 3), we derive a height for MMC 18 of 130.9 cm. (4 ft. VA in.), which is in close agreement with our first determination but may not be quite as accurate since Pearson's technique demands reading from a graph, and the regres- sion lines on his graph were based upon a general study of several populations of Old World pygmies. However, for the Mbuti he had data only on the two skeletons studied by Flower ( 1888), and of these the male is quite fragmentary. A further problem is that at the time Pearson was devising his procedures for determining statures from the lengths of various long bones, he had no adequate 430 FIELDIANA: GEOLOGY, VOLUME 33 To gain perspective on the stature of MMC 18 as compared to other Mbuti pygmies, we have summarized (table 3) the researches of Schebesta (1938) and Gusinde (1956) on this topic, adding also their determinations of weights on smaller series. Schebesta 's publi- cation we found to be of particular use since he evaluated the results of prior investigators and, from his own experiences, discounted some of them on the basis either that they were measuring popu- lations which contained in part genes from taller neighbors or by contrast had chosen to measure only the shorter individuals, even from groups all of whom were definitely pygmies. Excluding such questionable cases, but including several other acceptable series in addition to his own data, Schebesta published summaries (range and mean) on the stature of 1,346 pygmies (table 3). Gusinde (1956) then gave mean only, for an additional 896. The means of the two series agree within a half-inch for the males and a quarter-inch for the females.1 Until additional postcranial material of gracile australopithecines produces evidence to the contrary, we will assume that the bodily proportions, as based on the great similarity of the innominates, are similar to those of the Mbuti pygmies, even though we note that Robinson ( 1972, p. 231) believed that the pelvis of Sts 14 was some- (Footnote 2 continued from p. 429) series of female skeletons of pygmies with which to work, so all of his formulae and regression-lines applied to males only. For a female pygmy, thus, he had to convert the length of a long bone into the length of the masculine equivalent, find the stature of that hypothetical male, and then reconvert that stature to a female's equivalent. Additionally, Pearson had determined to his own satisfaction that Old World pygmies in general usually have longer femora relative to stature than do populations of more normal size, and he thus drew a different regression line for this relationship in pygmies than for taller populations. However, we have shown that the Mbuti (Aka) female described by Flower (1888) did not have such a relatively longer femur, and both Gusinde ( 1956, p. 119) and Turnbull ( 1965a) have mentioned their subjective impressions that in the Mbuti the legs are short and the arms long relative to the length of the trunk. Thus, at least for the Mbuti, Pearson's formula will generate a determination of stature somewhat greater than it should be, but the discrepancy is not as great as that produced by the use of the technique of Trotter and Gleser( 1952). 'Even within the Mbuti, stature varies between groups. Thus the Efe, as meas- ured by Schebesta (1938, p. 228) have the males averaging 2 cm. shorter than the males of the Basua and the mean height of the Efe females was nearly 3 cm. less than that of Basua females. Consequently, the ranges and means for stature re- ported by different investigators have varied somewhat depending upon what numbers of different groups were measured. CO .c ,g N . N a to 6 «>o> 60 O CC to o • t- £> "S JO eO eo — CO m* lO — i 1-1 $4 t-H ■<}< oo eo oo «i" ■1 3 a 1 00 co OS 5 J £ S 1 "2 > 'C 5 w CO as CO £ CO CD H < i E-i 60 oo a -u 3 .O s I g .s s e tO CD 60 C u to lit 1 CS d K da 1— 1 ^ '~H o ° « • oo CN . .9 w £ i - CD E o 1-H <*• 60 C a K i> to to CN to 1 .9 s .5 60 tO O CN 60 O _i to CO to § CO £ id O o to 1 CN N O 00 00 & eo I K co ■«r H os Oi e~ 00 to - oo co - 00 M CO to eo eo i 2 CD a)" Oi ■3 ^ i 2 CD " CD "9 •O 1 .0 c CD CD X * CO 3 a '55 9 0 431 432 FIELDIANA: GEOLOGY, VOLUME 33 what large for the remainder of her skeleton. On the basis, then, of the data here presented, we suggest that the stature of Sts 14 was most probably close to 129 cm. (4 ft. 2% in. ). Robinson (1972, pp. 121-125) estimated the bicondylar length of the left femur of Sts 14 as being between 30.0 and 32.0 cm., with a most probable length of 31.0 cm., as based on a study of the existing piece of shaft and neck, of the impressions of parts of the neck and head in the original matrix, and of two pieces of distal femur from Sterkfontein (pieces which would seemingly have conformed in general size and proportions to the proximal piece of Sts 14). As noted above, the estimate of total femoral length of Sts 14 reached by us is 35.0 cm. Since bicondylar length and total length vary from each other by less than 1 per cent, the difference is insignificant under the present circumstances for an estimated femoral length of 30.0-32.0 cm., as determined by Robinson, and we will ignore this particular factor and treat his bicondylar estimate as if it were total length. Robinson (1972, pp. 231-232) has discussed the height of Sts 14, as based on a study of various of the bones, but considered a calcu- lation based on his estimate of femoral length of 31 cm. to be the most accurate. Using Trotter and Gleser's (1952) formula for American black and white females together, he arrived at a figure of 4 ft. 3 in. Our duplication of this calculation yielded instead 4 ft. 3% in. ; we assume that Robinson probably rounded this off to the nearest inch. The difference is not great enough to fuss over the arithmetic, but both figures are erroneous since, as mentioned before, the regression formula of Trotter and Gleser produces an- swers which are too high for pygmies or small australopithecines. If we recalculate the stature of Sts 14, using Robinson's estimated length of 31 cm. for the femur, on the basis that the femur was 27.15 per cent of the height, we get a stature of only 114.2 cm. (3 ft. 9 in.); doing the same with Pearson's formula for living pygmies, we get 114.4 cm. (3 ft. 9V16 in.). If Robinson's estimate of a femoral length, 31.0 cm. for Sts 14, is accurate, then these latter figures for stature are, we believe, more probable than anything in the range of 4 ft. 3 in. to 4 ft. 4 in. However, if our estimate of femoral length of approximately 35 cm. is accurate, as we believe, and if we consider the femoral length to have been 27.15 per cent of the total height, then a stature near 130 cm. (approximately 4 ft. 3 in.) is most prob- ably correct. In other words, Robinson got the right answer, but for a couple of wrong reasons ; we think his estimate of femoral length REED & FALK: STERKFONTEIN 14 433 Table 4. Comparisons of ratios of measurements on innominate bones and fe- moral lengths of Sts 14, MMC 18, and a group of modern human skeletons {a = standard deviation; V = coefficient of variation). See p. 436 for discussion. AB/Fem m Specimens n Range Mean Range of Mean ±_ la Range of Meani2o Skeletons, Department of Anthropology, UICC Pygmy MMC 18 Gracile australopithe- cine Sts 14 11 38.2-34.0% 35.6% 1.23% 3.5 37.2-34.7% 38.4-33.5% 34.7% 34.5% Specimens n Range ac/ Fern m Mean a Range of Mean ±_ la Range of Meani2o* Skeletons, Department of Anthropology, UICC 11 21.9-18.9% 20.55% 0.925% 4.4 21.5-19.6% 22.5-18.7% Pygmy MMC 18 1 20.6% Gracile australopithe- cine Sts 14 21.4% was too short, but then he used a formula which, for pygmies, com- pensated by producing an erroneously high answer. We can ourselves simply consider the maximum length of the femur of Sts 14 as having been identical to that of MMC 18, at 35.0 cm., in which case her stature would have been the same as, or closely similar to, that of the pygmy, 128.9 cm. (4 ft. 23A in.).1 'The details of this paper would have been much simpler, and the conclusions undoubtedly sounder, if the stature of each individual Mbuti pygmy from which a skeleton has been derived had been determined before maceration. Such skeletons for which we have seen data number 10: two by Flower (1888), five by Matiegka (1938; the measurements of four of these five skeletons were separately published by Matiegka and Maly, 1938), and one each by Shrubsall (1902), Broek (1940), and Toerian (1954). Of these 10, only that of the female Aka has had the stature pub- lished, and that was determined by Flower from the skeleton alone. This important omission of lack of recorded stature is particularly difficult to understand in the case of Shrubsall's study, since a photograph of the particular pygmy was taken before his death and published with the study; presumably the subsequent history of skeletonization was not planned at the time of the photography. o i B £ 3 GO tf CO w o w O *3 pa .o ,Q CO .9 * oj CU i< C J3 1 1 * c q o <£ 05 J3 «« o 8 o *» > c o co c A o K "S « 6 a p m £2 -c a *J CO >i *_; TJ m a CM TJ T3 CU Cj | .9 1 6 _ £ 2 CO o rt CO A bo co.£ i 1 S s CO 5 A o « o X) * c >> « c CO co 6 i - J* g Tf O co V1 eo t~ =3 » £§ CO .9 2 ta 6. 8 S 22 •3 S o G Q) fl) 'A\ a So 175 l- J TJ > >-> CO !> ^, -u 2 o 2 SJ3 J5 2 3 *> P >.s a , . CD £ u *J CN T t- co g •—* ' CO cu U'H - !3 ri 3 -0 ^. <41 CU o « — b£ _ 3 C co eo cu cu £ £ » Q, ,cu J3 1! J» CO H cj O 2 2 J3 —item CO i— I ■~- — O (-, 1- CU CU CU '•3 ■§ CO c CO CO C In cu CU 5 p o E-1 3 o 73 s o> ,2 cu £ 2^ o u S -2 Si -° *» c co co -3 .2 *» C o S a "^ S § c S "3 "" JJ j 3 S Is* ^^ g -o "^ K co t— co cn CO 3 I I be © S -" 3 o 43 .a o PQ .2 c « co O a d a of CN t> t~ Oi 2 CO co • bo cu E S * fc — ' cu cu ■-* s I- ft 2 2 •: G A a •a o 5 a +» S q s 15 _ 8 * ^5 "cj co r> - -8 "S N. 2 2 S j= CO S S J m "O "3 c c 2 .2 « A ■u ja cu "3 "So cu 1 1 > CO — I »H o -^ 9 * P A 52 B « 3 o 6 .2B5 3 «a «m > c ° . C co 2-2-2 » §!^ *a OS B O CO ^3 £ x> A r "g bo c 0 w j= «oi • bo .. ** A >- co 53 6^ S « IB cu >-> „•« ftp ■^ CO c '-'go ^ w u co -B "2 9 « fl ® « m r h « o h CO cu > co CO co co C O O J5 05 % .. c 3 c ■« CU QJ . o. c o '** "1 3 — E T3 S cu - 8 I-1 vtS CU QJ ^" >^ ■* E H n « n fl 9 -u o •-| (X! co co co 434 CO 0) -4J 2 c c O ~0 co -2 s Jrf 5r eo | CO CO *"2 cj 1c eo g O jo CO 1 *> CO 73 3 a g CO <£ co G eo § S o CN t> OS 1— 1 c 3 -2 -w cu £ 1 J3 S gh; the tion dif sed. 1 O o CO C j3 1 £ "5 c £ 1c JS « CO O CO £ E o V CO > 'C -a 1 8 T-l CO *-> ! I o CU «*h 2 o 1-3 ii 3 5 £ Q.2 suit is erroni be used for upon which i o co u a co CO co co M ca >> c _2 £ CO co cu - J* IE h 3 o u CJ £ 1 >> £ U ca <*. *j B Si X 4-i ca cj 'E >> a C £ c 0) 'S CO ^o £ 9 X '-3 ca < go c _c _0) """ | CO la -*' CO 3 ca +5 £ a -4Jt CD -a cu ca IS 4J CO CO CO JS <*— a is c c c i G .c -a jz T3 N t> -H o t> 04 CO a X. CM T3 c J2 ca PQ 09 a CO 1 9 Z 2 co a CO .£ CN w 0) E 3 ca ■4-> K o o o 8 CO rf re 1-H 1— 1 ■^ 1-H .5 o £ <*- 1*4 c J= *J bX c S OJ hi C 3 Sj ki a E CJ T3 CO CO 1 CO CO rH CO ao o 3 s c O -a CO ca £ CJ C> to eo o -C £ c c c c co .9 CO o c c . E £ p in eo o JS tc c JS la t_ o E T3 co *j CO J w CO CO CO -u 0 >~ 3 c o -a 0 en CO E u q in en O -c £ o c .£ o CO .a CO o a o 9 CO a '•3 3 s c CO CO $3 O CO CO CO ■si CJ H >§ eo . o of S « ll bo co CO E CO CO CO *o 3 c o s CO ca E CJ p eo "o -c E o c _c "S CO B <^ o c c X ca c '£ 3 C ca -T CO •J O CO CO CO ,2 E Is . o of S ^> 2 "c3 b o 2 S bO CO CO 'c p cU C > a CO CO c Jj o CJ a. co cu El, c Jj c a cs cc 435 436 FIELDIANA: GEOLOGY, VOLUME 33 Such a simplistic approach does not, however, allow for possible individual variation within a population, so we have attempted to determine such variability by measuring the distances AB, AC, and the maximum femoral length (hereafter abbreviated to Femm) on each of 11 disarticulated modern human skeletons in the Depart- ment of Anthropology, University of Illinois at Chicago Circle. For these skeletons, the ratios AB/Femm and Ac/Femm were then calcu- lated, and in Table 4 the results are presented in comparison with the same ratios for MMC 18 and Sts 14. For MMC 18 and Sts 14, both ratios (AB/Femm and Ac/Femm) fall within the range, or immediately at the border of one side of that range, of the mean ±_ la. Thus, for both pygmy and Sts 18, the ratios of the proportions of the pelvis (as measured) to the maxi- mum length of the femur ( as known for the pygmy and as assumed therefrom for Sts 18) are in no way disproportionate to those of modern skeletons, and we assert with confidence, thus, that the femoral length of Sts 18 was at, or very close to, 35.0 cm., and that her height was at, or very close to, 4 ft. 23A in. The relatively small values of the Coefficients of Variation (V) of the ratios AB/Femm and Ac/Femm for the modern skeletons (table 4) strengthen the implica- tions of our conclusions.1 As McHenry (1974) has stated, the weight of an individual is a difficult factor to evaluate, given only a few broken bones. To help with an understanding of the relationship between height and weight, some comparisons with Mbuti pygmies (table 3) and with non-pygmoid populations may be useful, even though not, of course, conclusive. At 128.9 cm. ( = approximately 130 cm., or 4 ft. 3 in. ), MCC 18 was among the shorter females of her group, and we estimate her weight to have been between 60 and 68 lbs. We cannot transfer this esti- mate of weight directly to Sts 14, however, since the latter's skele- ton has a small acetabulum (pi. I) and a correspondingly small femoral head, as well as a slender femoral shaft, delicate sacro-iliac articulation, and small lumbar vertebrae (Robinson, 1972, p. 232). By all evidence, however, she was adult at time of death. Hopefully, 'We are not indulging ourselves in a completely circular argument. True, we originally guessed that the femoral length of Sts 14 would be close to that of the pygmy MMC 18, but then checked this opinion against innominate/femoral ratios of a series of modern normal-sized skeletons and found our original conclusion to be sustained. There does exist a logic behind the concept of the total morphologic pattern. REED & FALK: STERKFONTEIN 14 437 comparison with modern growing girls of the same height will yield more useful data. Kark (1953) published on the height and weight of 338 Bantu- speaking (mostly Zulu) girls aged 8-15 living in Durban, South Africa. These girls were better-fed, heavier, and taller in general than were girls of matched ages from a rural area near Durban. Most of the girls in the urban group attained a height of 129 cm. (our estimate for the stature of Sts 14) during their tenth to eleventh years (ages 9-10), at which time their mean weight was 29.5 kg. (65 lb. 1 oz.). Presumably the girls from the country, even though older when attaining 129 cm., would have been lighter in weight. A somewhat similar study of girls of African descent from poorer families in Jamaica (Ashcroft et al., 1966) showed that a height of 129 cm. was typically achieved during the ninth or tenth year of life (ages 8-9), at which time the median weight was 25.5 kg. (56 lb. 3 oz.). Girls of similar ancestry but from more affluent families reached the same height about a year younger but weighed 26.6 kg. (581b. 10 oz.) at the time. Sts 14 was not a growing child but was a young adult, and thus presumably equivalent in maturation to a modern girl in her early twenties; Sts 14 may, thus, have added a touch of fat to her frame, but at the same time, with her delicate skeleton, was presumably more lightly built than modern girls of her height. With all of these indeterminate variables an exact weight for Sts 14 cannot be calcu- lated; Robinson (1972, p. 232) guessed 40-60 lb. (18.2-27.3 kg.) as the weight for a female gracile australopithecine with a height of 129 cm.; considering the weights of living girls of that height, as well as our estimate ( 60-68 lb. ) of the weight of a female pygmy of the same height, we suggest a figure within the upper half of Robin- son's range, possibly at about 24-25 kg. (approximately 53-55 lbs. ) REFERENCES Allbrook, D. 1961. The estimation of stature in British and East African males based on tibial and ulnar bone lengths. Jour. Forensic Med., 8, pp. 15-28. Ashcroft, M. T., P. Heneage, and H. G. Lovell 1966. Heights and weights of Jamaican school children of various ethnic groups. Amer. Jour. Phys. Anthropol., n.s. 24, pp. 35-44. Broek, A. J.P. v. d. 1940. Das skelett einer weiblichen Efe-Pygmae. Z. Morphol. Anthropol., 38, pp. 121-169. 438 FIELDIANA: GEOLOGY, VOLUME 33 Broom, R. 1950. Finding the missing link. Watts and Co., London, vi + 111 pp. Broom, R. and J.T.Robinson 1947. Further remains of the Sterkfontein ape-man, Plesianthropus. Nature, 160, pp. 430-431. 1950. Further evidence of the structure of the Sterkfontein ape-man, Plesian- thropus. Transvaal Mus. Mem. no. 4, pp. 1-83. Clarke, R. J. and F. C. Howell 1972. Affinities of the Swartkrans 847 hominid cranium. Amer. Jour. Phys. An- thropol, n.s. 37, pp. 319-335. Dart, R. A. 1925. Australopithecus africanus: The man-ape of South Africa. Nature, 115, pp. 195-199. Flower, H. F. 1888. Description of two skeletons of Akkas, a pygmy race from central Africa. Jour. Royal Anthropol. Inst. Gt. Brit. Ireland, 18, pp. 3-19. GUSINDE, M. 1956. Die Twiden: Pygmaen und Pygmoide im tropischen Afrika. Arch. Vdlker- kunde Wien, 3, pp. 1-174. Kark.E. 1953. The growth and nutritional status of Bantu girls in Durban. So. Afr. Jour. Med. Sci., 18, pp. 109-124. Lovejoy, C. O. and K. G. Heiple 1970. A reconstruction of the femur of Australopithecus africanus. Amer. Jour. Phys. Anthropol, n.s. 32, pp. 33-40. Matiegka, J. 1938. Das Skelett der Ituri-Bambuti, pp. 321-353 In Die Bambuti-Pygmaen vom Ituri. I. Band: Geschichte, Geographie, Umwelt, Demographie und Anthro- pologic der Ituri-Bambuti (belgisch Kongo), by P. Schebesta. Mem. Inst. Roy. Colon. Beige, In-4°, Section Sci. Morales et Politiques. Matiegka, J. and J. Maly 1938. Etude de quatre squelettes de pygmees centre-africains du Bassin de l'lturi. 1' Anthropologic 48, pp. 237-248, 521-538. McHenry, H. M. 1974. How large were the australopithecines? Amer. Jour. Phys. Anthropol., n.s. 40, pp. 329-340. Pearson, K. 1899. Mathematical contributions to the theory of evolution. — V. On the recon- struction of the stature of prehistoric races. Phil. Trans. Royal Soc. London, Ser. A, 192, pp. 169-244. Robinson, J. T. 1967. Variation and the taxonomy of the early hominids. Evol. Biol., 1, pp. 69-100. 1972. Early hominid posture and locomotion. University of Chicago Press, Chi- cago, xi + 361 pp. REED & FALK: STERKFONTEIN 14 439 Sampson, C. G. 1974. The stone age archaeology of southern Africa. Academic Press, New York, xiv + 518 pp. SCHEBESTA, P. 1938. Die Bambuti-Pygmaen vom Ituri. I. Band: Geschichte, Geographie, Um- welt, Demographie und Anthropologic der Ituri-Bambuti (belgisch Kongo). Mem. Inst. Roy. Colon. Beige, In-4°, Section Sci. Morales et Politiques. xviii + 438 pp. Shrubsall, F. C. 1902. Remarks on the skeleton of a Bambute pygmy from the Semliki Forest, Uganda borderland, pp. 559-565 In Johnston, Harry, The Uganda Protectorate, Vol. II, Hutchinson and Co., London. Tobias, P. V. 1973. New developments in hominid paleontology in South and East Africa. Ann. Rev. Anthropol., 2, pp. 31 1-334. TOERIEN.M. J. 1954. The skeleton of a central African pygmy and its affinities. So. Afr. Jour. Med. Sci., 19, pp. 97-104. Trotter, M. and G. C. Gleser 1952. Estimation of stature from long bones of American whites and negroes. Amer. J. Phys. Anthropol., n.s. 10, pp. 463-514. 1958. A re-evaluation of estimation of stature based on measurements of stature taken during life and of long bones after death. Amer. Jour. Phys. Anthropol., n.s. 16, pp. 79-123. TURNBULL.C. M. 1965a. The Mbuti pygmies of the Congo, pp. 281-317 In Gibbs, J., ed., Peoples of Africa, Holt, Rinehart and Winston, Inc., New York. 1965b. Wayward servants: The two worlds of the African pygmies. Natural His- tory Press, Garden City, N.Y. xiv + 390 pp. ADDENDUM When writing this report, we overlooked a publication by Walker (1973), in which he stated that the femoral fragment of Sts 14 was so thoroughly fragmented and that the head had been so completely rebuilt in plaster that he " could see no way of reasonably esti- mating any measurements or angles and, short of gaining the im- pression that this is the smallest known femur of Australopithecus, very few useful points of comparison can be made." Additionally, Walker determined that the two distal femoral fragments from Sterkfontein were from distincly larger individuals than Sts 14, which fact casts doubt on those prior estimates which had depended upon the use of those distal fragments for determining the length of the femur of Sts 14. 440 FIELDIANA: GEOLOGY, VOLUME 33 However, Walker then concluded (in spite of his own statements regarding the lack of evidence to be derived from a study of the femoral fragment of Sts 14), that if intact that femur may have measured no more than 25 cm. in length, which is even less than the "preferred length" of 27.6 cm. of Lovejoy and Heiple ( 1970). Both of these estimates, as we have indicated in the present publication, seem to us to be too short, considering the size of the associated in- nominate. In a more recent publication, McHenry (1976) has compared the two surviving vertebrae of Sts 14 with their counterparts in 43 non- obese specimens from the Terry Collection of modern human skele- tons, for which body weights are known, and derived an estimate of 27.6 kg. (61 lb.) for the weight of Sts 14. This estimate is approxi- mately 13 per cent higher than ours, but, considering the different methods used, and the possible errors in each method, the agree- ment is relatively good. However, the sample of specimens used from the Terry Collection was half males, which could be assumed to average heavier than females of the same height, and the sample included individuals from 18 to 50 years of age, which probably re- sulted in a higher ratio of weight to height than if a sample of young adult females had been available. Considering these factors, per- haps McHenry 's estimate of 27.6 kg. is a bit high. McHenry, H. M. 1976. Early hominid body weight and encephalization. Amer. Jour. Phys. An- thropol.,45,pp. 77-84. Walker, A. 1973. New Australopithecus femora from East Rudolf, Kenya. Jour. Human Evol., 2, pp. 545-555. TOt ••*«r